University of Groningen
EEG neurofeedback for executive functions in children with neurodevelopmental challenges
Landes, Jasmin K; Reid, Corinne L; Arns, Martijn; Badcock, Nicholas A; Ros, Tomas;
Enriquez Geppert, Stefanie; Bulsara, Max K; Brini, Stefano; Rabipour, Sheida; Mason,
Mimma
Published in:
Cochrane Database of Systematic Reviews
DOI:
10.1002/14651858.CD012890
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2017
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Landes, J. K., Reid, C. L., Arns, M., Badcock, N. A., Ros, T., Enriquez Geppert, S., Bulsara, M. K., Brini, S.,
Rabipour, S., Mason, M., Birbaumer, N., Gouldthorp, B., & Anderson, M. (2017). EEG neurofeedback for
executive functions in children with neurodevelopmental challenges. Cochrane Database of Systematic
Reviews, 2017(12), 1-22. [ CD012890]. https://doi.org/10.1002/14651858.CD012890
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Cochrane
Database of Systematic Reviews
EEG neurofeedback for executive functions in children with
neurodevelopmental challenges (Protocol)
Landes JK, Reid CL, Arns M, Badcock NA, Ros T, Enriquez-Geppert S, Bulsara MK, Brini S,
Rabipour S, Mason M, Birbaumer N, Gouldthorp B, Anderson M
Landes JK, Reid CL, Arns M, Badcock NA, Ros T, Enriquez-Geppert S, Bulsara MK, Brini S, Rabipour S, Mason M, Birbaumer N, Gouldthorp B, Anderson M.
EEG neurofeedback for executive functions in children with neurodevelopmental challenges. Cochrane Database of Systematic Reviews 2017, Issue 12. Art. No.: CD012890.
DOI: 10.1002/14651858.CD012890. www.cochranelibrary.com
T A B L E O F C O N T E N T S 1 HEADER . . . . 1 ABSTRACT . . . . 1 BACKGROUND . . . . Figure 1. . . 4 Figure 2. . . 5 7 OBJECTIVES . . . . 7 METHODS . . . . 12 ACKNOWLEDGEMENTS . . . . 13 REFERENCES . . . . 20 APPENDICES . . . . 21 CONTRIBUTIONS OF AUTHORS . . . . 21 DECLARATIONS OF INTEREST . . . . 21 SOURCES OF SUPPORT . . . .
[Intervention Protocol]
EEG neurofeedback for executive functions in children with
neurodevelopmental challenges
Jasmin K Landes1, Corinne L Reid2, Martijn Arns3,4,5, Nicholas A Badcock6,7,8, Tomas Ros9, Stefanie Enriquez-Geppert10, Max K Bulsara11, Stefano Brini1,12,13, Sheida Rabipour14, Mimma Mason15, Niels Birbaumer16,17, Bethanie Gouldthorp1, Mike Anderson1 1School of Psychology and Exercise Science, Murdoch University, Perth, Australia.2School of Health in Social Science, The University of Edinburgh, Edinburgh, UK.3Research Institute Brainclinics, Nijmegen, Netherlands.4neuroCare Group, Munich, Germany. 5Department of Experimental Psychology, Utrecht University, Utrecht, Netherlands.6Department of Cognitive Science, Macquarie University, Sydney, Australia.7Perception in Action Research Centre, Macquarie University, North Ryde, Australia.8ARC Centre of Excellence in Cognition and its Disorders, Macquarie University, Sydney, Australia.9Department of Neuroscience, University of Geneva, Geneva, Switzerland.10Clinical and Developmental Neuropsychology, University of Groningen, Groningen, Netherlands. 11Institute for Health Research, The University of Notre Dame Australia, Fremantle, Australia.12Department of Psychology and Speech-Language Pathology, University of Turku, Turku, Finland.13Turku Brain and Mind Center, Turku, Finland.14School of Psychology, University of Ottawa, Ottawa, Canada.15Pearson Australia Group Pty Ltd, Sydney, Australia.16Institute of Medical Psychology and Behavioral Neurobiology, Eberhard-Karls-University Tuebingen, Tuebingen, Germany.17WYSS Center for Bio and Neuroengineering, Geneva, Switzerland
Contact address: Jasmin K Landes, School of Psychology and Exercise Science, Murdoch University, Project KIDS, Building 190, 90 South Street, Perth, Western Australia, 6150, Australia.J.Landes@murdoch.edu.au,jasmin.landes@gmail.com.
Editorial group: Cochrane Developmental, Psychosocial and Learning Problems Group. Publication status and date: New, published in Issue 12, 2017.
Citation: Landes JK, Reid CL, Arns M, Badcock NA, Ros T, Enriquez-Geppert S, Bulsara MK, Brini S, Rabipour S, Mason M,
Birbaumer N, Gouldthorp B, Anderson M. EEG neurofeedback for executive functions in children with neurodevelopmental challenges.
Cochrane Database of Systematic Reviews 2017, Issue 12. Art. No.: CD012890. DOI: 10.1002/14651858.CD012890.
Copyright © 2017 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
A B S T R A C T
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess the effectiveness of EEG neurofeedback as treatment for inhibition and updating problems in children facing neurodevelop-mental challenges.
B A C K G R O U N D
Neurodevelopmental disorders encompass a range of conditions, each with cognitive challenges that become apparent during child-hood. Historically, these have been conceptualised as various dis-tinct disorders on the basis of clinical phenotype and the classifica-tion of symptom clusters in theDiagnostic and Statistical Manual of Mental Disorders (DSM) Fourth Edition Text Revision
(DSM-IV-TR) orFifth Edition (DSM-5), or in the International Clas-sification of Diseases, 10th Revision (ICD-10) (APA 2000; APA 2013;WHO 1993). Neurodevelopmental disorders include con-ditions such as intellectual disability; autism; Asperger syndrome; fetal alcohol spectrum disorder (FAS); fragile X syndrome; Down syndrome; and attention deficit hyperactivity disorder (ADHD) (Bishop 2008;Bishop 2010). However, existing classifications for
neurodevelopmental disorders are not mutually exclusive, nor are these taxonomic systems intended to speak to causation - this presents a significant limitation when treatment options are con-sidered.
Recent development of the Research Domain Criteria (RDoC) has encouraged scientists to undertake transdiagnostic reanalysis of disorders in the interests of identifying shared causal pathways and consequently informing more effective prophylactic or cura-tive responses, or both (Coghill 2015). At the same time, emerg-ing literature on neural circuitry is illuminatemerg-ing development of circuits that mediate core complex cognitive processes and be-haviours in ways that cut across these diagnostic groups (Glahn 2016). Rather than distinct disorders, it is hypothesised that differ-ent phenotypes emerge as a result of the complex interplay of envi-ronment and biology within these neural circuits (Cuthbert 2010; Rutter 2010). For example, common cognitive functions that are implicated in a range of mental illnesses as well as in intellec-tual and behavioural disorders include reward response, emotional regulation, extinction, working memory, and response inhibition (Insel 2010a;Leckman 2010). In cognitive terms, these processes are summarised under the umbrella term ‘executive functions’ or cognitive control (Davidson 2006;Garvey 2016). They form the point of transdiagnostic intersect for several neurodevelopmental disorders (Doyle 2015;Insel 2010b).
Inhibition and updating dominate the executive function litera-ture and are the cognitive functions that have received the most attention as key functions of executive control (Miyake 2000;St. Clair-Thompson 2006). As such, they form a secure conceptual platform for this review. Inhibition refers to the capacity to in-hibit task-irrelevant distractors and to resolve conflict in pursuit of a particular goal (Barkley 1997;Hughes 2002;Miyake 2000; Padmanabhan 2015;Sergeant 2000). Inhibition is implicated in tasks such as the Stroop or Flanker task. Updating refers to the functioning of working memory and is implicated in coding of new information and, accordingly, in revising the information that is currently held in working memory. This function is targeted by tasks such as letter or digit memory and the Rey Auditory Ver-bal Learning Task (Kane 2003;Miyake 2000). Electroencephalo-graphic (EEG) neurofeedback training provided to target inhibi-tion and updating is described in the literature.
In sum, EEG neurofeedback treatment for neurodevelopmental disorders targeting the circuitry for inhibition or updating provides hope for prevention and remediation and will serve as the focus of this review.
Description of the condition
A considerable proportion of the population is affected by neu-rodevelopmental disorders including the following.
1. Autism. It is estimated that 1 in 160 children worldwide has a diagnosis of autism, which equates to more than 7.6 million disability-adjusted life-years and 0.3% of the global burden of disease (WHO 2013;WHO 2016a).
2. ADHD. It is estimated that 39 million people are affected by ADHD globally (WHO 2013). American reports suggest that approximately 11% of children between 4 and 17 years of age (6.4 million) are affected (Visser 2014). Australian figures suggest that 7.4% (298,000) of 4- to 17-year-olds who had a mental disorder between 2013 and 2014 suffered from ADHD (Lawrence 2015). Between 5 and 14 years of age, an estimated 3.4% of total years (1800 years) is lived in ill health or with disability, making ADHD the eighth leading cause of non-fatal loss of health for children in this age group in Australia (AIHW 2011).
3. Intellectual disabilities. An international meta-analysis in 29 countries indicated that, on average, 10.37 individuals among every 1000 people are affected (Maulik 2011). This is the seventh leading cause of non-fatal loss of health for children between birth and five years of age in Australia, with an estimated 4.3% of total years (700 years) lived in ill health or with disability (AIHW 2011).
4. Down syndrome, FAS disorder, and fragile X syndrome. These conditions have received little attention in statistical accounts; therefore, epidemiological data on these specific intellectual disorders are limited to prevalence rates. The incidence rate for Down syndrome is estimated to be around 1 in 1000 to 1 in 1100 live births worldwide (WHO 2016b). Western Australia reported estimates of FAS disorder of 0.4 per 1000 live births for the total population between 2000 and 2004 (Bower 2007). Roozen indicated that between 1990 and 2005 the reported occurence of FAS disorder in Canada, Italy, and the United States was in between 30.52 and 47.13 per 1000, and, in South Africa, the prevalence of FAS disorder is particularly high, at 113.22 per 1000 (Roozen 2016). In the absence of life expectancy data for fragile X syndrome, and given the strong genetic component involved in development of this disorder, prevalence rates are expected to be the same across all age groups (Brown 2010). Leykin reported that numbers of Australian persons with fragile X syndrome were expected to range between 1362 and 4309 for a full mutation with intellectual disability, and Brown anticipated numbers of 13,466 and 87,137 with a permutation (Brown 2010;Leykin 2009). Crawford estimated that 1 in 3717 individuals of European descent is affected by this condition (Crawford 2002). Youings projected that 1 in 5530 persons in the United Kingdom would receive a diagnosis of fragile X syndrome, and, most recently, Coffee anticipated that fragile X syndrome would occur in 1 in 5161 males in the United States (Coffee 2009;Youings 2000). In sum, neurodevelopmental disorders exact a significant toll on individuals, families, and communities. Gaining an
or remediation should be seen as a priority.
Difficulties with cognitive control are evident in the neurocog-nitive profiles of many individuals with different neurodevelop-mental disorders and are implicated in the behavioural and emo-tional presentation of affected children (Happé 2006;Pennington 1996). For example, problems related to inhibition and updat-ing are present to varyupdat-ing degrees. The neurocognitive profiles of children with fragile X syndrome and ADHD most often indicate problems with inhibitory control, which overlap with clinical fea-tures of impulsivity and hyperactivity (Bari 2013;Happé 2006; Hooper 2008;Knox 2012;Oosterlaan 2005). Difficulties with updating are implicated in Down syndrome, fragile X syndrome, and intellectual disability, all of which have been associated with limited ability to hold, manipulate, and process incoming infor-mation (Daunhauer 2014;Hartman 2010;Hooper 2008). Executive functioning, the central mechanism required for cogni-tive control, refers to the ability of the cognicogni-tive system to co-ordi-nate internal processes (e.g. perceptual selection and maintenance of contextual information) in pursuit of performance of specific tasks (e.g. reading a book) (Botvinick 2001). Although phenotypic presentations of difficulties in this area can be diverse, their impact on functioning generally aggregates and worsens as an individual gets older (Goldstein 2010;Masten 2005). Thus, when treatments for executive functioning challenges are considered, patient age is critical.
Executive functions develop rapidly during childhood (from about the age of six years) and adolescence (e.g.Anderson 2002). Not only does executive control typically improve during this time, but the nature of these functions is more fully differentiated (Brydges 2012;Brydges 2014a;Shing 2010). The growing neu-roplasticity literature informs us that integral to the maturation of a child’s nervous system are sensitive (but not necessarily crit-ical) periods for development (Davis 2009;Happé 2014;Heim 2012;Knudsen 2004;Newport 2001;Perani 2003;Wachs 2014; Weber-Fox 1996). During these sensitive periods, the brain is par-ticularly susceptible to change through experience, with poten-tial for diminished remediation in adulthood. Therefore, consid-ering treatment possibilities, such as EEG neurofeedback, during emerging stages of executive functions provides hope for reme-diating long-term dysfunction (Sonuga-Barke 2010). This devel-opmental period accounts for the choice of age groups included in the present review. We will focus on children and adolescents
between 6 and 18 years of age. As executive dysfunction plays a role in various disorders, review authors will not discriminate be-tween disorders. Instead, we will focus on core executive functions targeted through EEG neurofeedback, specifically, inhibition and updating.
Description of the intervention
Cognitive-behavioural therapy, behavioural intervention, medi-cal treatment, or a combination of at least two of these is cur-rently employed to manage symptoms of neurodevelopmental disorders (Ageranioti-Bélanger 2012;Fabiano 2009;Hsia 2014; Moskowitz 2011;Murawski 2015;Narzisi 2014;Reichow 2011; Scheifes 2013;Weston 2016). Conclusions regarding the effective-ness of non-invasive symptomatic treatment approaches are often limited by methodological weaknesses, such as lack of method-ological rigour or lack of randomisation during group allocation, and future research is needed to further evaluate treatment op-tions (Ozonoff 1998;Reid 2015;Walters 2016). The current al-ternative to non-invasive approaches is pharmacological interven-tion. The pharmacological treatment pathway serves as a popular means of symptom control, particularly for children with ADHD (Banaschewski 2006; Faraone 2010; Scheifes 2013). Although pharmacological interventions may be deemed a moderately ef-fective treatment option, side effects (e.g. headaches, dizziness, re-duced appetite, growth restriction), lack of certainty around po-tential long-term risks, reappearance of symptoms after discontin-uation of treatment, and non-response to medication have sparked the search for non-invasive long-term treatments that can be pro-vided without negative consequences (Graham 2011; Heinrich 2007;Jensen 2007;Murray 2008).
Technical advances have seen the development of EEG neuro-feedback as a promising, non-pharmacological mode of interven-tion that can be used to help train, prevent or remediate partic-ipants’ cognitive impairment at the source. EEG neurofeedback is commonly conceptualised as computer game-based training of awareness or control of cognitive state that can be achieved by providing participants with real-time feedback on their own brain states (Figure 1). It is thought that participants can learn to modify or control targeted brain-state activity, inducing neural plasticity, which leads to improved self-regulation in daily activities.
Figure 1. Neurofeedback intervention loop. This figure is adapted from Bagdasaryan and Le Van Quyen. It depicts a simplified overview of neurofeedback that is delivered via electroencephalography (EEG) (Bagdasaryan 2013). During the neurofeedback session, the individual’s brain signal is acquired through the EEG equipment (A, B). The software processes the incoming brain signal and provides information about the
degree of alignment between the participant’s real-time brain activity and predetermined training goal parameters (C, D). This information is presented to the participant as visual or auditory feedback in real time, to continuously update the participant about modulation of his or her own brain activity (E; Bagdasaryan 2013;
Huster 2014).
EEG neurofeedback is not dependent on complex verbal instruc-tions; therefore, this brain-training intervention can be effectively implemented cross-culturally and in groups with language and communication impairment. It is designed to be embedded in a game format, which offers face validity as a treatment for chil-dren. Currently, neither implementation of this approach in the community nor training of healthcare providers is monitored by an accredited organising body. Instead, implementation of EEG neurofeedback is based on the personal preferences of providers and consumers, which makes a systematic review of the evidence base for this treatment a critical task.
EEG neurofeedback training comprises a range of elements. A fundamental technical component of this intervention is the tech-nology that is used to monitor the degree of alignment between the participant’s brain activity and the pre-set goal parameter.
Var-ious brain-imaging techniques have been utilised for the inter-vention, such as near-infrared spectroscopy (NIRS), functional magnetic resonance imaging (fMRI), and electroencephalography (EEG) (Egner 2001;Marx 2015;Scharnowski 2015). Experimen-tal research to appraise various brain-imaging techniques is ongo-ing; however, to date, no one technique has been identified as the superior method for obtaining neurofeedback, nor is compelling evidence available to support the use of one technique over an-other. Therefore, we argue that to get the most accurate picture of changes in brain activity, such as those required in a micro-analysis of learning, a measure that can capture the most fine-grained changes in milliseconds rather than seconds is preferable (Sauseng 2008). Because it offers the highest temporal resolution of all known techniques, EEG is the only technology that meets
this criterion. Additionally, EEG is less costly and is more widely used than alternatives (Figure 2). Its use is also more feasible than, for example, fMRI when healthcare providers are working with children, as the experience of being inside an MRI machine can be unsettling and may disrupt optimal task completion.
Figure 2. Publication rates between 2006 and 2016 of journal articles examining EEG-, fMRI-, and NIRS-neurofeedback, as indexed by Scopus.FootnotesEEG: electroencephalogram; fMRI: functional magnetic
resonance imaging; NIRS: near-infrared spectroscopy.
EEG is a non-invasive method that is used to measure the brain’s electrical activity along the scalp via sensors or electrodes. During setup of EEG neurofeedback, these electrodes are placed on the head of the participant, and training goal parameters define fea-tures of the EEG signal that are to be trained (i.e. EEG frequency); the number of electrodes; the placement location of electrodes on the scalp; and the number, duration, and frequency of sessions (e.g. to decrease theta/alpha activity and increase beta activity, one electrode at Cz for 12 sessions overall, split into three 45-minute sessions per week; to increase sensorimotor rhythm and suppress
theta activity, individualised training at, for example, C3 and C4, for 30 sessions overall, split into two 20-minute sessions per week; and to increase sensorimotor rhythm and suppress theta activity, electrode placement at C4 for 40 sessions overall, split into 20 meetings of two 60-minute sessions each with a short break in between (Arnold 2013;Lansbergen 2011;Perreau-Linck 2010)). Once EEG neurofeedback equipment is set up, the participant’s neural variability is recorded and analysed in real time, and the degree of alignment between set parameters and the person’s
neu-ral activity is communicated in the form of visual or auditory feedback. Visual feedback may consist of a blurry picture of an animated character. Only when the participant’s ongoing neu-ral variability matches goal parameters would the picture become clear. Alternative visual feedback may be received in the form of virtual reality tasks, animated games, waveforms, or graphs (e.g. Linden 1996). Auditory feedback may comprise tones that change in volume, pitch, or duration as recorded activity fluctuates (Egner 2002).
The elaborate setup of the intervention poses interesting chal-lenges for comparison or control groups in EEG neurofeedback research. As the technology is integral to EEG neurofeedback, the most rigourous and the only form of placebo for which blinding is possible is sham EEG neurofeedback. During sham EEG neuro-feedback, the participant receives feedback unrelated to his or her own performance but based on pre-recorded or artificial EEG ac-tivity. Apart from the technical elements that make it challenging to find a placebo or carry out blinding, participants unsuccessfully attempt to learn to modulate their (fictitious) neural activity, of-ten resulting in poor compliance or frustration. Therefore, rather than subjecting participants to a blinding trial, EEG neurofeed-back research frequently implements alternative comparators such as conventional treatments (i.e. active, non-invasive control trials, for example, behavioural management interventions) and wait-list controls, by which participants wait for their active treatment intervention (Sonuga-Barke 2013).
Uncertainty in the literature pertains to the measurable effect of EEG neurofeedback, as well as to effects of technology, target fre-quency, electrode location, feedback type, number of sessions, and session duration on the efficacy of the intervention. Each of these intervention components forms a critical part of the intervention. In theory, any changes in the composition of these parts can influ-ence the efficacy of EEG neurofeedback and may constitute dis-tinct therapeutic approaches. In the absence of an evidence base for EEG neurofeedback, clinical practice currently operates on the basis of literature that has produced favourable EEG neurofeed-back outcomes in the past (e.g.Arnold 2013;Lubar 1995a). Each component needs further research to explore its influence on the intervention process. With a growing number of EEG neurofeed-back studies and approaches, it is not clear which frequency, elec-trode location, number of sessions, technology, or feedback type provides the best EEG neurofeedback therapy for children with executive functioning problems. As a starting point, we will look at the current literature and will investigate the fundamental ef-fectiveness of EEG neurofeedback.
How the intervention might work
Traditionally, researchers have conceptualised the neurofeedback loop as a learning process, which follows behavioural learning mechanisms of operant conditioning. Operant conditioning stud-ies (e.g. Schedules of Reinforcement) have shown that targeted participant behaviour can be regulated by providing positive
rein-forcement immediately after the targeted behaviour occurs (Ferster 1957). It is of great importance that the relationship between the behaviour and the reinforcement is clear to the learner. As such, the timing of presentation of the reinforcer is crucial, as even small delays (as little as a second) can decrease the strength of the as-sociation between the reinforcer and the target behaviour that is to be reinforced (Skinner 1958). Use of EEG in the neurofeed-back paradigm has enabled researchers to seek more immediate and more secure associations via measurable aspects of behaviours and reinforcers, such as by targeting the specific oscillations that underlie clinical symptoms like impulsivity or hyperactivity as the independent variable (Gevensleben 2012). It is therefore conceiv-able that the capacity for EEG neurofeedback to provide sub-second feedback may make it especially efficient as an approach to behaviour modification and brain plasticity, as compared with mental regulation unassisted by feedback (Bai 2014;Beatty 1974). Nevertheless, it should be noted that this potential advantage of high temporal resolution of EEG over other neurofeedback tech-niques has not been demonstrated.
Progressing technology has enabled researchers to seek clues in an attempt to describe EEG neurofeedback mechanisms from a biological viewpoint. As mentioned earlier, during EEG neuro-feedback, the participant is provided with feedback about differ-ences between target parameters and their actual neural activity. In theory, through this feedback, the participant can learn to mod-ulate brain activity towards the target parameter. Fundamental to this step is the idea that, during training, the participant learns to memorise the neural or behavioural state at the time of the reward, which facilitates reproduction of this same pattern in the future. Currently, the mechanisms that underpin this learning process have not been fully illuminated. However, principles of neuroplas-ticity may provide further clues to the causal pathway.
Neuroplasticity refers to the unique ability of the brain to grow neurons and to alter neural connections in response to experi-ences (Siegel 2010). Imaging studies have indicated, for exam-ple, how training in activities such as music, exercise, or medita-tion can have a lasting impact on brain structure or funcmedita-tion, or both (Vance 2010;Zatorre 2013). This finding highlights the fact that repeated, activity-dependent experiences can have a lasting impact on the brain (Ganguly 2013). Converging evidence sug-gests that reinforcing a particular oscillatory pattern through EEG neurofeedback training increases the likelihood that the same pat-tern will be reproduced more easily in the future (Lubar 1995b; Ros 2010). For the beta rhythm, for example, this effect is robust enough to be detected up to three years after EEG neurofeed-back training (Engelbregt 2016). This supports the fundamental premise of EEG neurofeedback that the brain can be conditioned to exhibit certain oscillatory patterns.
Theoretically, this phenomenon might be explained by a com-bination of previously established plasticity mechanisms, such as associative and homeostatic forms (Ros 2014). The principle of associative (i.e. Hebbian) plasticity suggests that “synapses that fire
together, wire together and synapses that fire apart, wire apart” (Knoblauch 2012). Neurophysiological evidence indicates that the amplitude of EEG oscillations is augmented by the number of neurons (or synaptic potentials) (Musall 2014). Therefore, with repeated training, the connections between neuron populations that are amplified or synchronised to create a particular oscillatory pattern would strengthen, facilitating generation of this pattern in the future. Common to all these theories is the hypothesis that modification of neural circuitry is possible and is likely to result in observable behavioural changes.
At a more technical level, during EEG neurofeedback, electrodes that are placed on the scalp measure the synchronised, rhythmi-cal fluctuations of lorhythmi-cal field potentials of groups of neurons, also known as neural oscillations. These oscillations arise from the exci-tatory postsynaptic potentials (EPSPs) of large groups of neurons, resulting in a measurable EEG signal at scalp level (Nunez 2000). Synchronised oscillatory activities are associated with cognitive abilities such as inhibition (for a review, see Klimesch), updating of working memory, and temporary maintenance of information in working memory, which suggests that neural oscillations are a fun-damental functional mechanism in cortical computation (Deiber 2007;Klimesch 2006;Sauseng 2010).
Why it is important to do this review
Executive functions play a critical role in everyday life. Perfor-mance of complex tasks, academic achievement, and later success in life are mediated by the development of executive functioning (Garavan 1999;Miyake 2000;St. Clair-Thompson 2006). EEG neurofeedback treatment provides hope for prevention and reme-diation of difficulties in the area of executive functioning for chil-dren with neurodevelopmental disorders. New EEG neurofeed-back protocols are continuously being tested to determine what most effectively reduces or remediates executive functioning dif-ficulties in neurodevelopmental disorders such as inhibition and updating (e.g.Kouijzer 2009). However, commercial use of EEG neurofeedback is currently outpacing the evidence base. The eco-nomic cost of this intervention is high for parents, but patient desperation is also high. A history of poorly researched intervtions for children (e.g. studies by Bishop and Stephenson) has en-couraged the profession to take greater accountability in establish-ing the effectiveness of new treatments as a matter of priority and professional ethics (Bishop 2007;Stephenson 2008). Use of EEG neurofeedback as an intervention for children with neurodevel-opmental problems has reached just such a critical juncture. It is imperative that the evidence base for these interventions is now put to the test.
This systematic review is the first of its kind and therefore will make a unique contribution to the EEG neurofeedback literature. It is preliminary to any further investigations of the cost-effective-ness or feasibility of this intervention. If the effectivecost-effective-ness of EEG neurofeedback is supported by this review, it is conceivable that
additional research will be conducted to identify its applicability to other mental health conditions or age groups.
O B J E C T I V E S
To assess the effectiveness of EEG neurofeedback as treatment for inhibition and updating problems in children facing neurodevel-opmental challenges.
M E T H O D S
Criteria for considering studies for this review
Types of studies
All randomised controlled trials (RCTs) (i.e. random allocation of participants to treatment, control, or follow-up groups) and quasi-RCTs (i.e. allocation of participants to intervention or control groups via date of birth, etc.).
Types of participants
Children or adolescents, or both, aged 6 to 18 years with execu-tive functioning difficulties in the domains of inhibition and up-dating. We will identify these neurodevelopmental challenges in the literature via the clinical diagnosis of a neurodevelopmental disorder such as intellectual disability, autism spectrum disorder (ASD), FAS disorder, fragile X syndrome, Down syndrome, and ADHD, as specified by theDiagnostic and Statistical Manual of Mental Disorders (DSM) Fourth Edition Text Revision
(DSM-IV-TR) orFifth Edition (DSM-5), or by the International Classifica-tion of Diseases, 10th Revision (ICD-10) (APA 2000;APA 2013; WHO 1993).
Excluded participants
We will exclude participants with severe brain damage, epilepsy, Tourette’s syndrome, or any condition in which the focus of neu-rofeedback intervention is not specific to executive functions (e.g. to remediate damage, seizures, or tics), as well as participants with non-neurodevelopmental comorbidities (e.g. depression). We will include participants with neurodevelopmental challenges as well as those with other mental health problems, as long as data provided for participants with neurodevelopmental challenges can be considered separately.
Types of interventions
EEG neurofeedback (also referred to as EEG biofeedback), regard-less of protocol (target frequency), feedback type (visual vs aural), and session number and duration, used as treatment for improv-ing levels of inhibition and updatimprov-ing (or both) in children with neurodevelopmental challenges. We will include studies adminis-tering EEG neurofeedback in combination with another interven-tion only when the cointerveninterven-tion is administered to both groups. For reasons outlined earlier in this protocol, control groups will include sham feedback (i.e. feedback that is unrelated to the partic-ipant’s neural activity at the time of intervention administration), treatment as usual (e.g. behavioural management intervention), and wait-list control (seeDescription of the intervention).
Types of outcome measures
Primary outcomes
1. Changes in participant EEG profiles (e.g. event-related potential (ERP), specifically, N2 for inhibition and P3 for updating (Brydges 2014b;Donchin 1988:Luck 2014;Polich 2007))
2. Changes in inhibition (e.g. Stroop Color and Word Test: Children’s Version) and changes in updating (e.g. Rey Auditory Verbal Learning Test) (Golden 2002;Schmidt 1996)
3. Adverse effects (e.g. Pittburgh Side Effects Rating Scale) (Pelham 1999) (It is important to note that we will consider only outcome assessments for which the outcome assessor was blinded.)
Secondary outcomes
1. Changes in behavioural performance (e.g. hyperactivity and impulsivity, as measured by self-report measures such as Conners-3 (Conners 2011)) (Again, we will consider only outcome assessments for which the outcome assessor was blinded.)
Search methods for identification of studies
Electronic searches
We will search the electronic databases and trials registers listed below, and will not limit our searches by language, date, or pub-lication type.
1. Cochrane Central Register of Controlled Trials (CENTRAL; current issue) in the Cochrane Library, which includes the Cochrane Developmental, Psychosocial and Learning Problems Specialised Register.
2. MEDLINE Ovid (1946 onwards).
3. MEDLINE In-Process and Other Non-Indexed Citations Ovid (current issue).
4. MEDLINE Epub Ahead of Print Ovid (current issue). 5. Embase Ovid (1974 onwards).
6. CINAHL Plus EBSCOhost (Cumulative Index to Nursing and Allied Health Literature; 1937 onwards).
7. PsycINFO Ovid (1806 onwards).
8. Science Citation Index - Expanded Web of Science (SCI-EXPANDED; 1970 onwards).
9. Social Sciences Citation Index Web of Science (SSCI; 1970 onwards).
10. Conference Proceedings Citation Index - Science Web of Science (CPCI-S; 1990 onwards).
11. Conference Proceedings Citation Index - Social Science & Humanities Web of Science (CPCI-SS&H; 1990 onwards).
12. Cochrane Database of Systematic Reviews (CDSR; current
issue), part of the Cochrane Library.
13. Database of Abstracts of Reviews of Effects (DARE; current issue), part of the Cochrane Library.
14. Epistemonikos (www.epistemonikos.org/nl/ advanced search).
15. WorldCat (OCLC) (www.worldcat.org/default.jsp). 16. ClinicalTrials.gov (www.clinicaltrials.gov).
17. World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP;apps.who.int/trialsearch). We will use the strategy provided inAppendix 1to search MED-LINE, and we will adapt it appropriately for searches of the other databases. When papers are published in a language other than English, we will contact the study author to request reports trans-lated into English.
Searching other resources
Grey literature
We will search the websites listed below for unpublished studies in this field.
1. The Association for Applied Psychophysiology and Biofeedback, Inc. (aapb.org).
2. The Biofeedback Federation of Europe (bfe.org). 3. EEG Spectrum International ( eegspectrum.com/ADHD-ADD).
4. International Society for Neurofeedback and Research ( isnr.org).
5. Applied Neuroscience Society of Australasia ( appliedneuroscience.org.au).
Reference lists
We will search the reference lists of all eligible studies included in this review for additional relevant studies that meet our inclusion criteria (seeCriteria for considering studies for this review).
Data collection and analysis
Selection of studies
Two review authors (JL and SB) will individually examine the titles and abstracts of records yielded by the search and will assess them against the inclusion criteria of this review (Criteria for considering studies for this review). For all studies that meet the inclusion cri-teria, or for which more information is needed to assess eligibility, we will obtain the full-text reports, and both review authors (JL and SB) will separately reassess these studies against the inclusion criteria. For full-text reports that are not written in the English language, or for data that are not available in the article, we will contact study authors for further information, to help us ascertain the eligibility of these studies for inclusion in the present review. We will record reasons for inclusion or exclusion of all studies sep-arately, and JL and SB will discuss discrepancies between views. When conflicting views cannot be reconciled, these review authors will consult the entire research team until a consensus is reached. Before any studies are selected, we will pilot-test application of the eligibility criteria. Finally, for maximum transparency during this selection process, we will complete a PRISMA flow diagram (Liberati 2009).
Data extraction and management
Two review authors (JL and SB) will independently extract and enter the following data from each study onto an electronic spread-sheet specifically designed for this Cochrane Review: participant details (age, gender, executive function problems); intervention details (number of sessions, session duration, follow-up, electrode location(s), frequency parameter, aural or visual feedback mode); study location; type of study (RCT, quasi-RCT); intervention pro-cedures (treatment allocation, blinding); type of control group (sham feedback, treatment as usual, wait-list control); and out-come measure data. Before any data are extracted, we will pilot test the application of spreadsheet categories, to ensure that rel-evant and comprehensive data are collected. We will resolve dis-agreements amongst ourselves that might occur during the data extraction process. When conflicting views cannot be reconciled, we will consult the entire research team until a consensus decision is reached.
Assessment of risk of bias in included studies
Two review authors (JL and SB) will independently assess the risk of bias of each included study, using the Cochrane ’Risk of bias’ tool (Higgins 2017). We will assess each study as having low, high, or unclear risk of bias in relation to each of the ’Risk of bias’ domains described below. JL and SB will record each rating separately and will discuss discrepancies between views. When conflicting views
cannot be reconciled, we will consult the entire research team until a consensus agreement is reached.
Cochrane ’Risk of bias’ tool
The domains described below form the ’Risk of bias’ assessments for RCTs and quasi-RCTs.
Sequence generation
We will describe the method used in each study to generate the participant allocation sequence and will assess whether this se-quence should have produced comparable participant groups. Review authors’ judgement: Is the participant allocation sequence truly random, and what is the resulting risk of allocation bias to experimental or control groups?
1. Low risk of bias: Study authors described the random component in the allocation sequence of participants (e.g. computer random number generator, random number table).
2. High risk of bias: Study authors described a non-random component in the allocation sequence of participants (e.g. allocation by date of birth or by judgement of the investigator).
3. Unclear risk of bias: The process of randomisation was not described in sufficient detail to permit a judgement of low or high risk of bias.
Allocation concealment
We will describe the measures that were used to conceal the allo-cation process from participants and from investigators and will determine whether this allocation to a particular group or training schedule could have been foreseen before, or during, participation by participants or investigators.
Review authors’ judgement: Is the participant allocation sequence truly concealed, and what is the resulting risk of allocation bias due to inadequate concealment?
1. Low risk of bias: The allocation procedure was adequately concealed from participants and investigators (e.g. telephone allocation).
2. High risk of bias: Participants or investigators could have foreseen their allocation (e.g. when allocation was based on the judgement of the clinician or on the date of birth of participants).
3. Unclear risk of bias: The allocation process was not described in sufficient detail to permit a judgement of low or high risk of bias.
Blinding of participants and personnel
We will describe all modes of blinding participants and staff from any knowledge of the intervention that a participant received. Review authors’ judgement: Are participants and personnel ade-quately blinded from any knowledge of the type of intervention
that participants received, and what is the resulting risk of perfor-mance bias?
1. Low risk of bias: Lack of blinding (no blinding or incomplete blinding) is present, but it is clear that this lack of blinding is unlikely to influence the outcome; or participants and personnel have been blinded, and it is unlikely that this blinding has been interrupted.
2. High risk of bias: Lack of blinding (no blinding or incomplete blinding) is present, and outcomes are likely to have been influenced by lack of blinding; or participants and personnel have been blinded, but it is likely that this blinding has been interrupted, which has influenced the outcome.
3. Unclear risk of bias: The blinding process was not described in sufficient detail to permit the judgement of low or high risk of bias, or this outcome was not addressed in the study. Owing to the learning component in EEG neurofeedback (seeDescription of the intervention), we expect that most studies will fall into this category.
Blinding of outcome assessment
We will describe all modes of blinding outcome assessors from any knowledge of the intervention that a participant received. Review authors’ judgement: Are outcome assessors adequately blinded from any knowledge of the type of intervention that par-ticipants received, and what is the resulting risk of detection bias? 1. Low risk of bias: The outcome assessment was not blinded, but it is clear that this lack of blinding is unlikely to influence the outcome measurement; or the outcome assessment has been blinded, and it is unlikely that this blinding has been interrupted, which has influenced outcome measurements.
2. High risk of bias: The outcome assessment was not blinded, and outcomes are likely to have been influenced by the lack of blinding, or the outcome assessment has been blinded, and it is likely that this blinding has been interrupted, which has influenced outcome measurements.
3. Unclear risk of bias: The blinding process was not described in sufficient detail to permit the judgement of low or high risk of bias, or this outcome was not addressed in the study. Owing to the learning component in EEG neurofeedback (seeDescription of the intervention), we expect that most studies will fall into this category.
Incomplete outcome data
We will describe the completeness of outcome data, including information on participant attrition, exclusions, re-inclusions for analyses, and participant numbers for each intervention group, as well as any withdrawals from study groups.
Review authors’ judgement: Are incomplete data handled ade-quately, and what is the resulting risk of attrition bias?
1. Low risk of bias: There is no indication of missing data; if data are missing, the same numbers of data points are missing across intervention groups; the data have been imputed suitably; or reasons for the missing data are unlikely to have influenced the outcome.
2. High risk of bias: Uneven numbers of data points are missing across intervention groups; data have been imputed through an unsuitable approach; or reasons for missing data are likely to have influenced the outcome.
3. Unclear risk of bias: Study authors did not provide sufficient information that permits a judgement of low or high risk of bias (e.g. no reasons for missing data provided), or this outcome was not addressed in the study.
Selective reporting
To assess any reporting bias, we will examine whether all prespec-ified outcomes have been reported. When this is not the case, we will contact researchers to ask for non-reported findings. Review authors’ judgement: Are there indications of selective out-come reporting, and what is the resulting risk of reporting bias?
1. Low risk of bias: Study protocol is available, and outcomes prespecified in the protocol have been reported; when the protocol is not available, it is clear that all expected outcomes have been reported.
2. High risk of bias: Not all of the outcomes prespecified in the protocol were reported; measurements were used that were not prespecified; outcomes were reported that were not prespecified; reporting of outcomes was incomplete; or study authors failed to include results for a particular outcome.
3. Unclear risk of bias: Study authors did not provide sufficient information to permit a judgement of low or high risk of bias. It is expected that most studies will fall into this category.
Allegiance bias
We will report any concerns of allegiance bias not otherwise cov-ered by above-mentioned components of the ’Risk of bias’ tool. Currently, no consensus has been reached on what constitutes an effective measurement of allegiance bias. The procedure most of-ten used to document and evaluate this type of bias was developed by Gaffan and colleagues (Gaffan 1995). However, in line with the critique provided by Leykin and colleagues, we will focus on eval-uation of the treatment protocol by its developers or, in this case, on sponsorship by neurofeedback equipment owners, to measure high, low, or unclear risk of allegiance bias (Leykin 2009).
1. Low risk of bias: no evidence that study authors developed the protocol; study was not sponsored by neurofeedback equipment owners.
2. High risk of bias: evidence that study authors developed the protocol; study was sponsored by neurofeedback equipment owners.
3. Unclear risk of bias: information provided points towards allegiance bias, but insufficient details prevent a judgement of whether low or high risk of bias is present; evidence of this bias is insufficient.
Measures of treatment effect
Continuous data
For continuous data, we will calculate the mean difference (MD), when possible (i.e. when the same outcome variables were assessed via the same measurement scale). When investigators assessed the same outcome variables through different modes of data collection (i.e. different scales, various scoring methods), we will calculate the standardised mean difference (SMD). We will present both the MD and the SMD with 95% confidence intervals (CIs).
Dichotomous data
For dichotomous data, we will compute the risk ratio (RR) for each outcome and the 95% CI to describe the probability that a particular outcome is going to occur.
Unit of analysis issues
Cluster-randomised trials
To our knowledge, no EEG neurofeedback studies have ran-domised groups or clusters of participants, rather than individuals; therefore, we do not expect to find any cluster-randomised trials during our search. Should cluster-randomised trials become avail-able in the future, we will assume that researchers have adjusted for clustering in their results. For trials that have not previously adjusted for clustering, we will attempt to calculate an estimate of the intracluster correlation coefficient (ICC) by using the for-mula provided in theCochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). If we are unable to calculate the ICC (owing to lack of information), we will request further infor-mation from study authors or will attempt to calculate the ICC using data from comparable studies or available resources. We will examine the impact of variation in ICCs via aSensitivity analysis and will discuss outcomes in the Discussion section of the review.
Cross-over trials
We do not anticipate identifying any cross-over trials. However, if we do, we will use data from the first period only, given the possibility of a carry-over effect, and lack of available information concerning the time taken for any intervention effects to fade or ’wash out’.
Studies with multiple interventions
We will combine all EEG groups and will conduct a simple, pair-wise comparison with all control groups. For participants who continue to receive medication, we will consider data only if par-ticipants in both intervention and control groups continue to re-ceive medication. We will conduct a sensitivity analysis to examine the potential effects of differences in participants’ medication or dosage, or both, on trial results (seeSensitivity analysis).
Multiple reports
When multiple reports describe the same study, we will take extra care to ensure that only independent findings are reported. If it is unclear whether reports include independent findings, we will contact the report authors to ask for clarification.
Dealing with missing data
We will record attrition and missing outcome data for each study and will contact study authors to request missing outcome data. When study authors do not provide data for missing summary statistics (e.g. standard deviations), we will base our calculations on other reported outcomes, when possible. When study authors do not provide data for missing specified outcomes, we will at-tempt to conduct an intention-to-treat (ITT) analysis by includ-ing participants randomised into a trial, irrespective of the group. If an ITT analysis is not possible, we will conduct an available case analysis using only participants whose outcome data are known. We will examine the impact of missing data on the main analyses via aSensitivity analysisand will discuss outcomes in the Discus-sion section of the review.
Assessment of heterogeneity
We will assess clinical and methodological heterogeneity by com-paring the effects of distribution of key participant traits (e.g. dis-tribution of sex, age, severity of executive functioning difficulties), protocol factors (e.g. target frequency, electrode location, feedback type, number of sessions, session duration), and trial factors (e.g. randomisation) on the efficacy of the intervention. We will em-ploy forest plots to identify any statistical heterogeneity (overlap of CIs) and will quantify this by computing I2and Chi2statistics (Deeks 2011). Although I2of 50% is a reasonable indication of heterogeneity, substantial heterogeneity will be clearly exemplified by I2of 65% (Section 9.5.2;Deeks 2011). The P value for the Chi2test must be less than 0.10. We will employ the Tau2statistic as a measure of between-study variability.
In the event of very high heterogeneity, we will identify studies that are contributing to high heterogeneity and will exclude them. If exclusion does not successfully remove the heterogeneity, we will not present outcomes of meta-analyses for this variable. We will transparently record all actions and reasons for exclusion.
Assessment of reporting biases
Before we include any studies, we will assess risk of allocation, de-tection, performance, attrition, and reporting biases, as outlined in theAssessment of risk of bias in included studiessection above. Additionally, when we include more than 10 studies, we will pre-pare funnel plots to assess publication bias. We will visually inspect these plots for skewness. When we find evidence of an asymmet-rical funnel plot, we will apply Egger’s test (Egger 1997).
Data synthesis
To conduct the meta-analysis, we will pool outcome data through Review Manager 5 (RevMan 5) (Review Manager 2014). Owing to the nature of our study design, we will consider the likelihood of heterogeneity in our data as high (e.g. data from varying EEG neurofeedback protocols, participants with different disorders and from different study designs).
Given the high probability of significant heterogeneity in our re-sults, we will apply a random-effects model. We will conduct subgroup analyses to systematically investigate heterogeneity (see Subgroup analysis and investigation of heterogeneity).
’Summary of findings’
We will create a ’Summary of findings’ table using a combina-tion of RevMan 5 (Review Manager 2014) and GRADEprofiler (GRADEpro; GRADEpro GDT 2015). In this table, we will present effects of EEG neurofeedback (1) on underlying ERPs of executive function performance, (2) on executive function perfor-mance as measured by psychometric tests, (3) in relation to overall well-being of participants with adverse effects recorded, and (4) on behavioural performance. Additionally, we will include the num-ber of participants and a rating of the quality of evidence based on GRADE criteria derived using GRADEpro (GRADEpro GDT 2015;Guyatt 2006;Schünemann 2006). Two review authors (JL and SB) will independently rate the quality of evidence according to one of four levels (high, moderate, low, or very low). For ex-ample, we will rate the quality of evidence from a RCT as high; however, presence of risk of bias (e.g. design limitations, limita-tions in the implementation of studies that are likely to introduce bias), indirectness of evidence (e.g. indirect effects on the popu-lation, intervention or control groups, or outcomes), imprecision (e.g. wide CIs due to small sample sizes), inconsistency of results (e.g. unexplained heterogeneity), and/or reporting bias (e.g. pub-lication bias; failure to report outcomes) may lower the GRADE rating. Both review authors will make notes to guide their judge-ments to ensure a transparent rating procedure.
Subgroup analysis and investigation of heterogeneity
Subgroup analyses are observational in nature, and any conclu-sions drawn are intended only to highlight potential areas of fu-ture research (Deeks 2011). When sufficient outcome data are available, we will carry out subgroup analyses and investigations of heterogeneity for each outcome (seeTypes of outcome measures), with consideration of the following points.
1. Investigation of the effectiveness of EEG neurofeedback as a function of frequency, session number, session duration, electrode location, or feedback type.
2. Investigation of the effectiveness of EEG neurofeedback as a function of the control group against which it is compared.
3. Investigation of the effectiveness of EEG neurofeedback as a function of the childhood disorder, as defined by diagnostic criteria (DSM-IV-TR; DSM-5; ICD-10) (APA 2000;APA 2013;WHO 1993).
4. Investigation of the effectiveness of EEG neurofeedback as a function of age.
5. Investigation of the interaction between intervention factors (e.g. session numbers) and dropout rates.
Sensitivity analysis
Our goal is to draw robust conclusions regarding the questions that we ask in this review. When methodological choices of individual studies or trial analyses might compromise the robustness of our conclusions, we will conduct sensitivity analyses. Specifically, we anticipate that we will be able to conduct sensitivity analyses for the situations listed below.
1. Comparison of variable findings from RCTs and quasi-RCTs.
2. Studies with high or unclear risk of bias as indicated by the ’Risk of bias’ assessment.
3. Studies with concurrent psychopharmacological treatment. 4. Variation in ICCs for analyses pertaining to cluster-randomised controlled trials.
5. Studies with missing data.
A C K N O W L E D G E M E N T S
We would like to thank the members of the Cochrane Develop-mental Psychosocial and Learning Problems Group, notably Dr Joanne Wilson, Professor Geraldine Macdonald, Margaret Ander-son, and Gemma O’Loughlin for processing our application and for providing helpful advice during writing of this protocol. We would also like to thank Margaret Solosy and Jean Coleman from Murdoch University for their support in drafting a search strategy.
R E F E R E N C E S
Additional references Ageranioti-Bélanger 2012
Ageranioti-Bélanger S, Brunet S, D’Anjou G, Tellier G, Boivin J, Gauthier M. Behaviour disorders in children with an intellectual disability.Paediatrics & Child Health 2012;
17(2):84–8. [PUBMED: PMC3299352] AIHW 2011
Australian Institute of Health and Welfare. Australian Burden of Disease Study: impact and cause of illness and death in Australia 2011. www.aihw.gov.au/WorkArea/ DownloadAsset.aspx?id=60129555476 (accessed 27 July 2016).
Anderson 2002
Anderson P. Assessment and development of executive function (EF) during childhood. Child Neuropsychology 2002;8(2):71–82. [DOI: 10.1076/chin.8.2.71.8724; PUBMED: 12638061
APA 2000
American Psychiatric Association.Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR). 4th Edition. Washington (DC): American Psychiatric Association, 2000. APA 2013
American Psychiatric Association.Diagnostic and Statistical Manual of Mental Disorders (DSM-5). 5th Edition. Washington (DC): American Psychiatric Association, 2013. Arnold 2013
Arnold LE, Lofthouse N, Hersch S, Pan X, Hurt E, Bates B, et al. EEG neurofeedback for ADHD: double-blind sham-controlled randomized pilot feasibility trial. Journal of Attention Disorders 2013;17(5):410–9. [DOI: 10.1177/1087054712446173; PMC3939717; PUBMED: 22617866
Bagdasaryan 2013
Bagdasaryan J, Le Van Quyen M. Experiencing your brain: neurofeedback as a new bridge between neuroscience and phenomenology. Frontiers in Human Neuroscience 2013;7(680):1–10. [DOI: 10.3389/fnhum.2013.00680; PMC3807564
Bai 2014
Bai O, Huang D, Fei DY, Kunz R. Effect of real-time cortical feedback in motor imagery-based mental practice training.NeuroRehabilitation 2014;34(2):355–63. [DOI: 10.3233/NRE-131039; PUBMED: 24401829
Banaschewski 2006
Banaschewski T, Coghill D, Santosh P, Zuddas A, Asherson P, Buitelaar J, et al. Long-acting medications for the hyperkinetic disorders. A systematic review and European treatment guideline.European Child & Adolescent Psychiatry 2006;15(8):476–95. [DOI: 10.1007/s00787-006-0549-0; PUBMED: 16680409
Bari 2013
Bari A, Robbins TW. Inhibition and impulsivity: behavioral and neural basis of response control. Progress
in Neurobiology 2013;108:44–79. [DOI: 10.1016/ j.pneurobio.2013.06.005; PUBMED: 23856628 Barkley 1997
Barkley RA. Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD.Psychological Bulletin 1997;121(1):65–94. [PUBMED: 9000892]
Beatty 1974
Beatty J, Greenberg A, Deibler WP, O’Hanlon JF. Operant control of occipital theta rhythm affects performance in a radar monitoring task. Science 1974;183(4127):871–3. [DOI: 10.1126/science.183.4127.871; 4810845 Bishop 2007
Bishop DVM. Curing dyslexia and attention-deficit hyperactivity disorder by training motor co-ordination: miracle or myth?. Journal of Paediatrics and Child Health 2007;43(10):653–5. [DOI: 10.1111/j.1440-1754.2007.01225.x; PMC2835859; UKMS28929 Bishop 2008
Bishop D, Rutter M. Chapter 3. Neurodevelopmental disorders: conceptual issues. In: Rutter D, Bishop DVM, Pine DS, Scott S, Stevenson J, Taylor E, et al. editor(s). Rutter’s Child and Adolescent Psychiatry. 5th Edition. Oxford (UK): Blackwell Publishing Ltd, 2008:32–41. [DOI: 10.1002/9781444300895.ch3
Bishop 2010
Bishop DVM. Which neurodevelopmental disorders get researched and why?. PLoS One 2010;5(11):e15112. [DOI: 10.1371/journal.pone.0015112; PMC2994844; PUBMED: 21152085
Botvinick 2001
Botvinick MM, Braver TS, Barch DM, Carter CS, Cohen JD. Conflict monitoring and cognitive control.Psychological Review 2001;108(3):624–52. [PUBMED: 11488380]
Bower 2007
Bower C, Rudy E, Callaghan A, Cosgrove P, Quick J, Nassar N.Report of the Birth Defects Registry of Western Australia 1980-2006. Perth (Western AU): Birth Defects Registry, 2007.
Brown 2010
Brown L.The Prevalence of Fragile X-Associated Disorders in Australia. Canberra (AU): NATSEM, University of Canberra, 2010.
Brydges 2012
Brydges CR, Reid CL, Fox AM, Anderson M. A unitary executive function predicts intelligence in children. Intelligence 2012;40(5):458–69. [DOI: dx.doi.org/ 10.1016/j.intell.2012.05.006
Brydges 2014a
Brydges CR, Fox AM, Reid CL, Anderson M. The differentiation of executive functions in middle and late childhood: a longitudinal latent-variable analysis. Intelligence 2014;47:34–43. [DOI: dx.doi.org/10.1016/ j.intell.2014.08.010
Brydges 2014b
Brydges CR, Fox AM, Reid CL, Anderson M. Predictive validity of the N2 and P3 ERP components to executive functioning in children: a latent-variable analysis.Frontiers in Human Neuroscience 2014;8:1–10. [DOI: 10.3389/ fnhum.2014.00080; PMC3929846
CNG 2013
Kerson C, Collaborative Neurofeedback Group. A proposed multisite double-blind randomized clinical trial of neurofeedback for ADHD: need, rationale, and strategy. Journal of Attention Disorders 2013;17(5):420–36. [DOI: 10.1177/1087054713482580; PUBMED: 23590978 Coffee 2009
Coffee B, Keith K, Albizua I, Malone T, Mowrey J, Sherman SL, et al. Incidence of fragile X syndrome by newborn screen for methylated FMR1 DNA.American Journal of Human Genetics 2009;85(4):503–14. [DOI: 10.1016/ j.ajhg.2009.09.007; PMC2756550; PUBMED: 19804849 Coghill 2015
Coghill D. Commentary: we’ve only just begun: unravelling the underlying genetics of neurodevelopmental disorders - a commentary on Kiser et al (2015). Journal of Child Psychology and Psychiatry 2015;56(3):296–8. [DOI: 10.1111/jcpp.12399; PUBMED: 25714739 Conners 2011
Conners CK, Pitkanen J, Rzepa SR. Conners Comprehensive Behavior Rating Scale. In: Kreutzer JS, DeLuca J, Caplan B editor(s). Encyclopedia of Clinical Neuropsychology. 3rd Edition. New York (NY): Springer, 2011:678–80. [link.springer.com/referenceworkentry/ 10.1007/978–0–387–79948–3˙1536]
Crawford 2002
Crawford DC, Meadows KL, Newman JL, Taft LF, Scott E, Leslie M, et al. Prevalence of the fragile X syndrome in African-Americans. American Journal of Medical Genetics 2002;110(3):226–33. [DOI: 10.1002/ajmg.10427; PUBMED: 12116230
Cuthbert 2010
Cuthbert BN, Insel TR. Toward new approaches to psychotic disorders: the NIMH research domain criteria project.Schizophrenia Bulletin 2010;36(6):1061–2. [DOI: 10.1093/schbul/sbq108; PMC2963043
Daunhauer 2014
Daunhauer LA, Fidler DJ, Hahn L, Will E, Lee NR, Hepburn S. Profiles of everyday executive functioning in young children with Down syndrome.American Journal on Intellectual and Developmental Disabilities 2014;119 (4):303–18. [DOI: 10.1352/1944-7558-119.4.303; PMC4512669; PUBMED: 25007296
Davidson 2006
Davidson MC, Amso D, Anderson LC, Diamond A. Development of cognitive control and executive functions from 4 to 13 years: evidence from
manipulations of memory, inhibition, and task switching. Neuropsychologia 2006;44(11):2037–78. [DOI: 10.1016/ j.neuropsychologia.2006.02.006
Davis 2009
Davis MC, Luecken L, Lemery-Chalfant K. Resilience in common life: introduction to the special issue. Journal of Personality 2009;77(6):1637–44. [DOI: 10.1111/ j.1467-6494.2009.00595.x; PUBMED: 19796066 Deeks 2011
Deeks JJ, Higgins JP, Altman DG, editor(s). Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JP, Greens S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated in March 2011). The Cochrane Collaboration, 2011. Available from handbook.cochrane.org.
Deiber 2007
Deiber MP, Missonnier P, Bertrand O, Gold G, Fazio-Costa L, Ibañez V, et al. Distinction between perceptual and attentional processing in working memory tasks: a study of phase-locked and induced oscillatory brain dynamics. Journal of Cognitive Neuroscience 2007;19(1):158–72. [DOI: 10.1162/jocn.2007.19.1.158; PUBMED: 17214572 Donchin 1988
Donchin E, Coles MGH. Is the P300 component a manifestation of context updating. Behavioural and Brain Sciences 1988;11(3):357–427. [DOI: 10.1017/ S0140525X00058027
Doyle 2015
Doyle AE. Commentary: insights from across diagnostic boundaries: ADHD in RDoC era - a commentary on Scerif and Baker (2015). Journal of Child Psychology and Psychiatry 2015;56(3):274–7. [DOI: 10.1111/jcpp.12401; PUBMED: 25714738
Egger 1997
Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315(7109):629–34. [DOI: dx.doi.org/10.1136/ bmj.315.7109.629; PMC2127453; PUBMED: 9310563 Egner 2001
Egner T, Gruzelier JH. Learned self-regulation of EEG frequency components affects attention and event-related brain potentials in humans. Neuroreport 2001;12(18): 4155–9. [PUBMED: 11742256]
Egner 2002
Egner T, Strawson E, Gruzelier JH. EEG signature and phenomenology of alpha/theta neurofeedback training versus mock feedback. Applied Psychophysiology and Biofeedback 2002;27(4):261–70. [PUBMED: 12557453]
Engelbregt 2016
Engelbregt HJ, Keeser D, van Eijk L, Suiker EM, Eichhorn D, Karch S, et al. Short and long-term effects on sham-controlled prefrontal EEG-neurofeedback training in healthy subjects. Clinical Neurophysiology 2016;127(4): 1931–7. [DOI: 10.1016/j.clinph.2016.01.004; PUBMED: 26971473
Fabiano 2009
Fabiano GA, Pelham WE Jr, Coles EK, Gnagy EM, Chronis-Tuscano A, O’Connor BC. A meta-analysis of behavioural treatments for attention-deficit/hyperactivity
disorder.Clinical Psychology Review 2009;29(2):129–40. [DOI: 10.1016/j.cpr.2008.11.001; PUBMED: 19131150 Faraone 2010
Faraone SV, Buitelaar J. Comparing the efficacy of stimulants for ADHD in children and adolescents using meta-analysis. European Child & Adolescent Psychiatry 2010;19(4):353–64. [DOI: 10.1007/s00787-009-0054-3; PUBMED: 19763664
Ferster 1957
Ferster CB, Skinner BF.Schedules of Reinforcement. New York (NY): Appleton-Century-Crofts, 1957.
Gaffan 1995
Gaffan EA, Tsaousis I, Kemp-Wheeler SM. Researcher allegiance and meta-analysis: the case of cognitive therapy for depression.Journal of Consulting and Clinical Psychology 1995;63(6):966–80. [DOI: 10.1037/0022-006X.63.6.966; PUBMED: 8543719
Ganguly 2013
Ganguly K, Poo MM. Activity-dependent neural plasticity from bench to bedside.Neuron 2013;80(3):729–41. [DOI: 10.1016/j.neuron.2013.10.028; PUBMED: 24183023 Garavan 1999
Garavan H, Ross TJ, Stein EA. Right hemispheric dominance of inhibitory control: an event-related functional MRI study.Proceedings of the National Academy of Sciences of the United States of America 1999;96(14): 8301–6. [PMC22229; PUBMED: 10393989] Garvey 2016
Garvey M, Avenevoli S, Anderson K. The National Institute of Mental Health research domain criteria and clinical research in child and adolescent psychiatry. Journal of the American Academy of Child & Adolescent Psychiatry 2016;55(2):93–8. [DOI: 10.1016/j.jaac.2015.11.002; PMC4724376; PUBMED: 26802775
Gevensleben 2012
Gevenleben H, Holl B, Albrecht B, Schlamp D, Kratz O, Struder P, et al. Is neurofeedback an efficacious treatment for ADHD? A randomised controlled clinical trial.Journal of Child Psychology and Psychiatry, and Allied Disciplines 2009;
50(7):780–9. [DOI: 10.1111/j.1469-7610.2008.02033.x; PUBMED: 19207632
Glahn 2016
Glahn DC, Knowles EEM, Pearlson GD. Genetics of cognitive control: implications for Nimh’s research domain criteria initiative.American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 2016;171(1):111–20. [DOI: 10.1002/ajmg.b.32345
Golden 2002
Golden CJ, Freshwater SM.Stroop Color and Word Test: Children’s Version. Wood Dale (IL): Stoelting Co, 2002.
Goldstein 2010
Goldstein S, Reynolds CR, editors. Handbook of Neurodevelopmental and Genetic Disorders in Children. 2nd Edition. New York (NY): The Guilford Press, 2010.
GRADEpro GDT 2015 [Computer program]
McMaster University (developed by Evidence Prime). GRADEpro GDT. Version accessed 10 May 2017. Hamilton (ON): McMaster University (developed by Evidence Prime), 2015.
Graham 2011
Graham J, Banaschewski T, Buitelaar J, Coghill D, Danckaerts M, Dittman RW, et al. European guidelines on managing adverse effects of medication for ADHD. European Child & Adolescent Psychiatry 2011;20(1):17–37. [DOI: 10.1007/s00787-010-0140-6; PMC3012210; PUBMED: 21042924
Guyatt 2006
Guyatt G, Gutterman D, Baumann MH, Addrizzo-Harris D, Hylek EM, Phillips B, et al. Grading strength of recommendations and quality of evidence in clinical guidelines: report from an American College of Chest Physicians task force. Chest 2006;129(1):174–81. [DOI: 10.1378/chest.129.1.174; PUBMED: 16424429 Happé 2006
Happé F, Booth R, Charlton R, Hughes C. Executive function deficits in autism spectrum disorders and attention-deficit/hyperactivity disorder: examining profiles across domains and ages.Brain and Cognition 2006;61(1): 25–39. [DOI: 10.1016/j.bandc.2006.03.004; PUBMED: 16682102
Happé 2014
Happé F, Frith U. Annual research review: towards a developmental neuroscience of atypical social cognition. Journal of Child Psychology and Psychiatry 2014;55(6): 553–77. [DOI: 10.1111/jcpp.12162
Hartman 2010
Hartman E, Houwen S, Scherder E, Visscher C. On the relationship between motor performance and executive functioning in children with intellectual disabilities.Journal of Intellectual Disability Research 2010;54(5):468–77. [DOI: 10.1111/j.1365-2788.2010.01284.x; PUBMED: 20537052
Heim 2012
Heim C, Binder EB. Current research trends in early life stress and depression: review of human studies on sensitive periods, gene-environment interactions and epigenetics. Experimental Neurology 2012;233(1):102–11. [DOI: 10.1016/j.expneurol.2011.10.032; PUBMED: 22101006 Heinrich 2007
Heinrich H, Gevensleben H, Strehl U. Annotation: neurofeedback - train your brain to train behaviour.Journal of Child Psychology and Psychiatry, and Allied Disciplines 2007;
48(1):3–16. [DOI: 10.1111/j.1469-7610.2006.01665.x; PUBMED: 17244266
Higgins 2011
Higgins JPT, Deeks JJ, Altman DG, editor(s). Chapter 16: Special topics in statistics. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 (updated March 2011).
